Images in Virtual Agent responses

[ayman_h]

In a Now Assist Virtual Agent implementation, if we have a Knowledge article with an image embedded within it, will the response in Virtual Agent ever include the image? I understand the context within image is not processed by the LLM. 

Only asking this as recently, I did notice an 'The images or media helped' option within the new granular feedback introduced in VA.

Regards,

Ayman

[MaryG]

I don’t believe an image embedded in a Knowledge article will automatically be included in the Now Assist-generated VA response.

Now Assist Q&A uses the article content to generate the response and provides citations back to the source article. Virtual Agent can display images, but that is handled separately through VA image response controls/topics rather than by pulling embedded images from the Knowledge article into the generated answer.

I wouldn’t read too much into the “images or media helped” granular feedback option either. Those feedback options are configurable and don’t necessarily indicate that Knowledge article images are being rendered by Now Assist.